Bitumen PG 76-16: Specification, the 92 Degree Span and Modification

Read the name as two design temperatures with a measurement standing behind each. PG 76-16 means the binder has been shown to satisfy the high-temperature criteria of AASHTO M320 at an average seven-day maximum pavement design temperature of 76 °C, and the low-temperature criteria at a minimum pavement design temperature of −16 °C. Both figures describe the asphalt surface at a stated reliability rather than the air above it, and both are demonstrated on binder that has been artificially aged before it is tested. Neither is assigned by inspection, deduced from a penetration result, or inherited from a refinery data sheet.

What separates this grade from the rest of the performance grade family is not either number taken alone. A 76 °C high grade is ordinary across the Gulf and the tropics. A −16 °C low grade is ordinary across the inland Middle East and North Africa. It is the distance between them that makes PG 76-16 a different commercial proposition from either of its neighbours, and that distance — the useful temperature interval, 92 °C — is what the middle of this page is about.

The grade name prints one test temperature and hides two

Everything a PG 76-16 report is built on can be worked out from the grade name before the report is opened, which is why a buyer with no laboratory can still audit a certificate line by line.

The shear tests sit at 76 °C. Nothing needs calculating on the hot side: the 76 in the name is the chamber setting for the dynamic shear rheometer, and two separate criteria are applied there. On original binder, G*/sin δ must reach 1.00 kPa. On rolling thin film oven residue — 85 minutes at 163 °C under continuous air to AASHTO T240 — the same parameter must reach 2.20 kPa. The step up to 2.20 is not the same question asked more harshly. Between the storage tank and the paver the binder passes through a hot plant and comes out stiffer, and the second criterion asks whether that stiffened material still refuses to move under a wheel through the summers that follow. In practice only the second one decides anything, because a binder that clears 2.20 kPa aged was never in danger of failing 1.00 kPa unaged.

The fatigue criterion sits at 34 °C. M320 tests it at neither grade temperature. It goes to the midpoint between them and then adds four degrees. Worked out for PG 76-16:

• (H + L) ÷ 2 + 4
• (76 + (−16)) ÷ 2 + 4
• 60 ÷ 2 + 4
• 30 + 4 = 34 °C

At 34 °C the pressure ageing vessel residue must show G*·sin δ of no more than 5000 kPa. Note that this parameter is a product rather than a quotient and that a low value is wanted: an aged binder that is simultaneously stiff and elastic stores strain energy under each wheel pass instead of dissipating it, and eventually gives that energy back as a crack. Because the figure tracks both halves of the grade name, it is also a quick authenticity check. PG 76-10 answers the same question at 37 °C and PG 76-22 at 31 °C, so a report headed PG 76-16 whose fatigue result was measured at 37 °C was graded against a −10 binder, not against yours.

The bending beam sits at −6 °C. The rule is L + 10, so −16 + 10 = −6 °C. This is the condition that is questioned most often, because ten degrees warmer than the grade name looks like a concession. It is not. A laboratory that ran the beam at −16 °C would have graded the sample against a different and much colder binder. The ten degrees are borrowed from time–temperature superposition: bitumen is viscoelastic, so how stiff it appears depends as much on how long a load is held as on how cold it is, and the two are interchangeable along a shift factor that can be measured on real binder. Thermal cracking is not instantaneous. An inland pavement radiates heat away for hours after sunset while tensile stress accumulates and the binder works to relax it, so an honest laboratory version at −16 °C would mean holding each specimen under load for about two hours, which is unusable as an acceptance test and leaves the beam barely deflecting far enough to read. The accepted equivalence is that 60 seconds of loading at L + 10 °C returns the same stiffness as roughly two hours of loading at L. The beam is therefore conditioned to −6 °C and read at the sixty-second mark, where creep stiffness S is capped at 300 MPa and the m-value must reach 0.300. The reading is taken at −6 °C; what it certifies is how the binder behaves at −16 °C across a real night.

On the cold side, read the m-value first

At −6 °C most paving binders are still a long way from 300 MPa, so creep stiffness usually passes with a wide margin and tells you very little. The m-value does not behave that way. Stress relaxation is the first property oxidative ageing attacks and the property that wax interferes with most directly, so on a −16 grade certificate the m-value is the number that separates binders that make the grade from binders that do not. A report that quotes S at −6 °C and omits the m-value has omitted the decisive result.

There is a second reason to read it first on this particular grade, and it is the point that connects the cold end to everything else on this page. Polymer buys most of its ground at the hot end. An elastomeric network raises the high-temperature grade substantially and improves relaxation at the cold end only modestly, so the low-temperature half of a PG 76-16 still has to be delivered by the base binder underneath the modifier. That is why a binder built to hold 92 °C is normally built on a softer base than one built to hold 86 °C at the same high grade, and why a supplier cannot simply add polymer to a PG 76-10 formulation and call the result PG 76-16.

76 is a traffic grade at least as often as it is a climate grade

Superpave practice adjusts the high-temperature grade for loading as well as for climate: one step of six degrees for slow transient loads, two steps for standing loads, with a further step considered for very high volumes. The adjustment is applied to the high-temperature grade only. The low grade is left where the site temperature record put it, because traffic does not change how cold the pavement gets at night.

That asymmetry is where most PG 76-16 tonnage comes from. A site whose climate produced PG 64-16 and whose traffic stands still — a signalised approach, a bus lane, a toll plaza, a container yard, a border crossing queue — is bumped two steps to PG 76-16 without its climatic high grade ever having been 76 °C. A site that produced PG 70-16 and carries slow climbing-lane traffic gets there in one step. In both cases the span widens purely because of traffic, and the binder has to be manufactured to a 92 °C interval all the same. The manufacturing problem does not care why the specification asked.

What the grade does not tell you

M320 is a modifier-blind standard. It does not require a polymer, does not name one, does not measure one and does not reward one. A binder either meets the criteria at the stated temperatures or it does not, and the certificate proves the grade rather than the route taken to reach it. It says nothing about crude source, nothing about modifier family or dosage, nothing about storage stability, and nothing about whether the material will still be homogeneous after a fortnight in a hot tank.

What the grade draws is a climatic and load envelope, and an envelope is all it draws. It is not a mix design, it says nothing about how the binder will bond to a particular aggregate, and it does not address moisture damage or stripping. The practical consequence for this grade in particular is blunt: modification is something you infer from the span, not something the standard writes on the certificate. That is precisely why it has to be asked about rather than assumed.

Two grade temperatures locate the binder; the interval sizes the problem

A pair of grade temperatures answers a question about place: where on the temperature axis this binder has been shown to work. The interval between them answers a different question, about process: how much had to be done to the material for one binder to cover that much ground. The two questions are independent, and reading only the first is how buyers end up comparing offers that are not comparable.

Take three grades with an identical high-temperature requirement. PG 76-10, PG 76-16 and PG 76-22 are all tested for rutting at exactly 76 °C, on original binder and on RTFOT residue, against exactly the same 1.00 kPa and 2.20 kPa limits. On the summer side they are the same product and the same test. Their intervals are 86, 92 and 98 °C, and those three numbers describe three quite different manufacturing problems: one comfortably inside neat capability, one at its outer edge, one beyond it. Nothing in the shared 76 hints at that, and a purchasing decision made on the high grade alone will not see it.

Now take the mirror case, grades with an identical interval. PG 82-10, PG 76-16, PG 70-22 and PG 64-28 all span 92 °C. They are comparably demanding to produce and they are not remotely interchangeable in service. A supplier who answers a PG 76-16 enquiry with PG 70-22 on the grounds that the span is the same has answered a question about production difficulty rather than about the climate and the traffic at your site. PG 70-22 is never tested at 76 °C, and the shear requirement at 76 °C is exactly what the project is buying.

So the two readings do different work, and a buyer needs both. The grade temperatures tell you whether the binder suits the site. The span tells you what kind of product you are actually being offered, how much had to be done to the material to produce it, and what will have to happen at the storage tank when it arrives.

Why the two ends fight each other

The interval is informative because the two criteria pull in opposite directions inside the same material. Stiffness and elastic response at 76 °C generally come with a higher asphaltene content and a stiffer maltene phase. Compliance and, above all, stress relaxation at −6 °C want the opposite. There is no free adjustment available, only a trade.

Every ordinary lever a refinery has works on both ends at once and in the same direction. A deeper cut on the vacuum column, a harder base residue, air blowing, prolonged hot storage: each of them lifts the high-temperature grade and gives ground at the low-temperature grade, usually by taking the m-value down first. Those levers slide the service window up the temperature axis. They do not widen it. A straight-run binder from a given crude occupies a window of roughly fixed width, and grade selection from a neat production stream is essentially a choice of where to put that window, not of how wide to make it.

That is the mechanical reason the interval, rather than the high grade, is the number that predicts whether a binder can be produced without a modifier. It is also why the interval is worth calculating on every performance grade offer that crosses a desk, including the ones that look unremarkable.

The working ceiling, and where 92 °C sits against it

The figure used across the industry is that a straight-run binder from a suitable crude will hold a useful temperature interval of up to roughly 92 °C, and that stretching both ends beyond about that point calls for modification. Treat it as a working rule rather than a specification limit: no standard states it, it is not a property of every crude, and a refinery with an unusually favourable feedstock will do better than one without. But it is the number the trade works to, and it places PG 76-16 in an unambiguous position. This grade does not sit inside the neat envelope with room to spare. It sits on its outer edge.

Set the common grades against that ceiling and the picture is clear:

• PG 64-16, span 80 °C — twelve degrees of headroom. Modification is neither implied nor justified, and an offer that presents the binder as modified deserves a question.
• PG 70-16, span 86 °C — six degrees. Frequently achievable neat from a suitable crude, which is why a one-step traffic bump from PG 64-16 usually stays in unmodified territory.
• PG 76-16, span 92 °C — none. This is the two-step bump, and it is the step that moves an order out of neat supply.
• PG 76-22, span 98 °C — well beyond it. Nobody argues about that one.

Effectively certain is the honest description of modification at 92 °C, and it is worth being precise about why the wording is not always. The rule of thumb has exceptions, and a genuinely exceptional crude processed carefully can be graded out at 92 °C without a modifier. What it cannot do is get there with margin. A neat binder that just reaches PG 76-16 has essentially nothing in hand at either end, and margin is what absorbs the ordinary events of a real supply chain: a hot week in the tank farm, a plant running at the top of its temperature range, a fraction of reclaimed asphalt in the blend that drags the effective low grade warmer, a batch drawn from the tail of a production run. A modified binder built to hold 92 °C comfortably and a neat binder that scrapes it are both correctly labelled PG 76-16 and they are not the same purchase. The continuous grade at both ends is what distinguishes them on paper, and it belongs on the test report.

What this means for a buyer reading an offer

If a supplier offers PG 76-16 and the offer says nothing about modification, that is a question to ask rather than a detail to skip. There are only three explanations worth considering, and separating them at the enquiry stage takes no more than a few questions.

• It is a modified binder and the offer simply did not say so. This is the commonest case and the least troubling on its own, but the omission is not neutral. A modified binder changes the storage arrangement, the agitation requirement, the heating regime, the mixing and compaction temperatures and the plant procedure at destination. A buyer who plans for an ordinary paving grade and receives a polymer modified one has acquired equipment obligations that were never allowed for.
• It is a genuinely exceptional neat binder. Possible, and easy to prove. A continuous grade stated at both ends, together with a full DSR and BBR set carrying the tank or batch reference of the cargo being loaded, settles the question in one page. A supplier who has such a binder will not find the request difficult.
• The grade designation is being asserted rather than measured. The commonest version of this is a penetration grade production stream described in performance grade language because the tender asked for performance grade language. There is no arithmetic that converts 60/70 into PG 76-16: a needle depth at 25 °C on virgin binder says nothing about how the material shears at 76 °C, nothing about how it relaxes at −6 °C, and nothing whatever about what twenty hours in a pressure ageing vessel leaves behind. A PG designation with no DSR and BBR data behind it is a claim, not evidence.

The questions that separate the three are short and none of them requires a laboratory to ask. Is the binder modified, and with which family of modifier. What is the continuous grade at each end. Was the grading done on the batch being loaded, or on production from some earlier period. Is there a separation result, and against what limit. What mixing and compaction temperatures does the supplier recommend, given that the equiviscous method a plant would normally use does not apply to a modified binder. The pattern of the answers is informative well before the numbers are.

One thing the span will not tell you

An interval is a statement about manufacturing difficulty. It is not a statement about quality, and it is not a statement about elastic response. Two binders can both hold 92 °C and behave differently under a standing wheel, because more than one route exists to a high-temperature grade and only some of them build an elastic network in the binder. That is a limitation of the M320 parameter itself rather than of the span argument, and it is the specific problem the multiple stress creep recovery test was written to solve. Both are dealt with below.

Sliding the window against widening it

Hold the distinction from the previous section in mind, because it separates the processes that can produce a PG 76-16 from the processes that cannot. Oxidation, deeper cutting and harder base residues move the whole service window upward: the binder gets better at 76 °C and worse at −6 °C in the same operation. A dispersed polymer network behaves differently. It contributes structure that carries load elastically at high service temperature while adding comparatively little rigidity at low service temperature, so the top of the window moves and the bottom does not follow it up nearly as far. The window gets wider.

That is the whole technical content of the phrase modification is effectively certain at 92 °C. It is not a marketing preference. It is a statement that the other available processes push both ends the same way.

The modifier families, and what each one actually buys

Elastomers are the family normally behind a wide span. Styrene-butadiene-styrene block copolymer is the dominant one, with styrene-butadiene rubber and latex products used in some markets. The polybutadiene mid-block swells by absorbing the maltene fraction of the binder, and above a critical concentration the swollen polymer forms a continuous network through the material. That network carries load at high service temperature and recovers when the wheel passes, which is why elastomeric modification raises the high-temperature grade and adds measurable elastic recovery at the same time. No standard fixes the dosage; published figures for paving binders sit in the low single digits as a percentage of binder mass, and the dosage that produces a continuous network depends on the base binder’s maltene chemistry as much as on the quantity added. Compatibility is a genuine constraint rather than a formality: not every base binder will carry a given polymer at a given dosage without separating in storage, which is the reason the separation test discussed below exists at all.

Plastomers — ethylene vinyl acetate, ethylene methyl acrylate, polyethylene — raise the high-temperature grade by stiffening rather than by building an elastic network. They can lift a high grade efficiently and they contribute little elastic recovery, so they do less for a wide span and nothing much for behaviour under a standing load. Where rutting under moving traffic is the entire problem they are a reasonable tool; where the specification is asking for 92 °C of span and elastic response together, they are a partial answer.

Crumb rubber from ground tyre rubber, blended by the wet process, swells in hot binder and lifts the high-temperature grade substantially while adding a measure of elasticity. The handling burden differs again: agitation is not optional, the blend has a working life during which it continues to react, and the material behaves differently in a pump and a spray bar from a polymer modified binder. It is a distinct product family with its own specification practice rather than an interchangeable alternative.

Chemical modifiers, most commonly polyphosphoric acid, raise the high-temperature grade by shifting the asphaltene and maltene balance inside the binder. They are used alone and as co-modifiers alongside polymer. The important point for a buyer is that they build no elastic network, so their contribution shows up in a creep compliance result and not in a recovery result. Some road authorities cap or restrict their use, and interactions with certain aggregates and with some amine anti-stripping additives are documented, so where a project specification is silent it is worth asking rather than assuming.

Why air blowing is not on that list

Air blowing raises the softening point and the high-temperature grade, and it is the simplest process available for making a binder harder. It also drives the m-value down, because oxidation attacks stress relaxation first. Applied to a paving binder aiming at a 92 °C span, it buys the top of the window at the direct expense of the bottom — exactly the wrong trade. Oxidised grades are industrial products for roofing, membranes and pipe coating, and the fact that an oxidised binder can be very stiff at 76 °C is not evidence that it could pass PG 76-16, because the bending beam at −6 °C is where it would fail.

The standard does not know any of this happened

M320 sets criteria at temperatures. It does not ask how they were met. Two binders can both certify as PG 76-16 — one carrying an elastomeric network, one lifted chemically or by a stiff base with a plastomer — and behave differently under a queue of standing trucks at a signalised intersection, which is very often the exact situation that caused PG 76 to be specified in the first place. The G*/sin δ parameter was developed on unmodified binders. It is measured in the linear viscoelastic region, at a strain no wheel ever applies, and it does not distinguish an elastic network from simple stiffness.

This is not a reason to distrust M320. It is a reason to specify more than the grade when the grade implies a modified binder, and to know which additional lines belong on the certificate.

What to add to the specification once modification is in play

• Separation of polymer — ASTM D7173. A sample is sealed in a tube, held vertically at 163 °C for 48 hours, cooled and cut, and the top and bottom sections are recovered and tested separately. The difference between them is reported, usually as a difference in softening point by ASTM D36 or as a difference in a DSR result. Note carefully what D7173 is: a practice that produces a number, not a specification that sets a pass mark. The acceptance limit belongs to your purchase specification, and an offer that quotes the method without a limit has quoted half the requirement.
• Elastic recovery — ASTM D6084. The ductilometer method, normally run at 25 °C on the binder or on recovered residue, still used in a great many national and agency specifications as the evidence that an elastomer is present and functioning.
• Multiple stress creep recovery — AASHTO T350. The modern route, described below, and the better one because it is run at the actual high-temperature grade on RTFOT residue rather than at an arbitrary ambient temperature on unaged material.
• A statement of the modifier family. Not the proprietary formulation, which no supplier will disclose, but whether the product is elastomeric, plastomeric, crumb rubber or chemically modified. That single line determines what the binder will do under standing load and how it must be stored.
• The supplier’s recommended mixing and compaction temperature ranges, in writing with the offer, because the calculation a plant would ordinarily perform does not apply to this material.

MSCR and AASHTO M332: the test that asks the right question

The multiple stress creep recovery test loads RTFOT residue in the dynamic shear rheometer at the high-temperature grade — 76 °C for this grade — for one second, then lets it recover for nine seconds, and repeats. Ten cycles are run at a stress of 0.1 kPa and ten more at 3.2 kPa. Two results come out of it. Non-recoverable creep compliance, written Jnr and reported in kPa⁻¹, is the strain that did not come back, per unit of applied stress: it is the direct measure of the permanent deformation the binder contributes to rutting. Percent recovery is the strain that did come back, and it is where an elastic network shows itself, because a binder stiffened without one recovers very little.

Together they answer the question G*/sin δ cannot. A binder can be stiff and non-elastic, or slightly less stiff and strongly elastic, and MSCR tells the two apart at the temperature the pavement actually reaches, under two stress levels rather than one, on aged material.

AASHTO M332 is the specification built on that test, and it changes the shape of the grade name. Instead of bumping the high-temperature grade for traffic, M332 keeps the climatic grade and adds a traffic designation letter — S, H, V or E — set by the Jnr result at 3.2 kPa, with a further requirement that the Jnr difference between the two stress levels does not exceed 75 %. Where an agency wants positive evidence of elastomeric modification, M332 provides a minimum percent recovery curve read against the Jnr value at 3.2 kPa, expressed as R3.2 ≥ 29.371 × (Jnr3.2)−0.2633 and applied where Jnr at 3.2 kPa is 2.0 kPa⁻¹ or lower.

Why this matters specifically to a PG 76-16 order

Recall where most PG 76-16 tonnage comes from: a −16 climate whose high-temperature grade was bumped two steps because the traffic stands still. M332 decouples those two things. The same project, specified to M332, keeps its climatic 64 and adds a V or an E, so the binder that would have been called PG 76-16 under M320 may be called PG 64V-16 or PG 64E-16 instead. The material is doing the same job. The specification is describing it more directly, by asking for elastic response and low permanent deformation at the temperature the site reaches, rather than for stiffness at a temperature the site may never see.

Two practical consequences follow. First, an M320 certificate contains no Jnr and no percent recovery, and neither can be derived from a G*/sin δ result — different residue conditioning is not the issue, but different stress levels, a different loading pattern and a fundamentally different parameter are. Establish which of the two standards governs the tender before the order is placed, not after the cargo has been graded. Second, when an offer against a PG 76-16 enquiry comes back quoting a traffic letter, that is not evasion; it is a supplier working to the other standard, and the correct response is to align the paperwork rather than to reject the material.

The pairing that defines the grade

A site that genuinely produces both halves of PG 76-16 from its own temperature record needs two things at once: summer insolation severe enough to drive a black surface to 76 °C, and a winter night cold enough to take that same surface to −16 °C. Those are not naturally companionable requirements, and the climates that produce both share a short list of features — dry air, clear skies, distance from a moderating sea, and usually elevation.

The physics is straightforward. Humidity and cloud act as a thermal blanket. Where they are absent, a surface absorbs solar radiation all day and then radiates it straight back to a clear sky at night with nothing to hold it in. Diurnal swings above 20 °C are commonly recorded in desert interiors, and asphalt exaggerates both ends: it is darker than the ground around it, so it runs hotter by day, and it is an efficient radiator, so at night it runs colder than the air above it. Elevation sharpens both effects. Thinner, drier air radiates heat away faster after sunset, so nights are colder, while a dark surface under a high-altitude sky receives more intense insolation, not less. High desert plateaux therefore widen the required span from both ends at once, which is exactly the condition a 92 °C interval describes.

Humidity is the variable that separates the two grades

On a humid coast the sea and the water vapour above it cap the day and floor the night. The pavement can be brutally hot in August and still not approach −16 °C in January, and frequently does not approach −10 °C either. That is PG 76-10 territory: the same 76 °C shear criterion on original binder and RTFOT residue, a low grade of −10 °C, a span of 86 °C, the bending beam run at 0 °C rather than −6 °C, and the fatigue criterion at 37 °C rather than 34 °C.

Note also where the 76 comes from in each case. On the humid coast it is almost never a climatic design temperature of 76 °C; it is a severe summer plus a traffic adjustment for container terminals, port approach roads, bus lanes and the queues at customs and weighbridges. Inland at elevation the 76 may be climatic or it may equally be a two-step bump applied to a 64 °C climate. The origin of the number does not change the manufacturing problem, but it changes the conversation with the engineer, because a traffic-derived 76 is a candidate for the M332 route described above and a climatic 76 is not.

Cooling rate, and why the −16 earns its place inland

Low-temperature cracking is not caused by a temperature. It is caused by a stress the binder could not relax quickly enough, which makes the rate of cooling as important as the depth of it. A clear desert night, with no cloud and little water vapour, produces some of the fastest surface cooling rates found anywhere. A binder with a marginal m-value has time to shed stress during a slow maritime cooling and very little time during a fast continental one.

That is why the failures the −16 prevents look different from northern thermal cracking. Instead of one catastrophic cracking event on the coldest night of a decade, inland desert pavements accumulate thermal fatigue: many moderate cool-down cycles each winter, each one loading and unloading the binder, gradually opening transverse cracks that then admit water into the structure. The absolute minimum temperature is unremarkable. The number of cycles is not, and it is the number of cycles that a coastal grade specified inland will not survive.

Which conditions choose which, stated plainly

• Inland, dry, at elevation, with a winter minimum pavement design temperature at or below −16 °C, and either a severe summer or standing and crawling traffic: PG 76-16. Both halves of the grade are being used.
• Humid coastal or tropical, with a winter that does not approach −16 °C, and heavy or standing traffic: PG 76-10. The extra six degrees of winter buys protection the site cannot use, and brings with it a wider span, a harder binder to manufacture and in practice a more heavily modified product.
• Inland at elevation, ordinary free-flowing highway traffic, no standing loads: PG 64-16 or PG 70-16. Specifying PG 76-16 here brings in a modified binder and its entire handling burden to solve a rutting problem the site does not have.
• Inland where the ground genuinely freezes and stays frozen: the low grade moves to −22. That is a 98 °C span, a different availability question and a different conversation with the supplier.

The error with the heaviest consequences is using a coastal grade inland. Specifying PG 76-10 for a high-altitude interior site exposes the pavement to precisely the thermal fatigue the −16 was there to prevent, and because the mechanism is cumulative the pavement looks sound for a season or two before the transverse cracking appears. The correct grade is the one the site’s own temperature record produces at the reliability the authority uses, not the one the region is assumed to need.

Region by region

• The Arabian interior at elevation. Central and northern Saudi Arabia and inland Oman. Extreme summer insolation with winter nights that reach or pass freezing away from the coast. Riyadh sits at roughly 600 m. The high grade is frequently traffic-derived on urban and industrial work.
• The Levant and northern Mesopotamia. The Jordanian plateau at around 800 m, inland Syria and northern Iraq. Hot continental summers, genuinely cold winter nights, and a wide daily range for most of the year — a textbook −16 climate.
• Central and eastern Anatolia. Interior Turkey combines intense summer sun at altitude, dry air and reliable sub-zero winter nights. Parts of the eastern uplands push past a −16 requirement into −22 territory, which is why site data rather than regional assumption has to settle it.
• The Maghreb interior. The Atlas foothills and the Moroccan and Algerian high plateaux, and the inland routes of Tunisia and Libya. The Mediterranean coastal strip a few hundred kilometres away is a different climate and usually a different grade.
• The interior southwest of the United States. Arizona, Nevada, New Mexico, west Texas and inland southern California sit squarely in this band, and PG 76-16 appears widely across that region’s specification landscape for heavy and slow-loaded pavement.
• The Mexican central plateau and comparable high interior basins. Elevation plus a dry season produces the same combination of severe surface heating and rapid nocturnal cooling.
• The plateaux of interior South and Central Asia. The Potohar and Balochistan plateaux, the drier interior basins further north, and comparable elevated interiors where summers punish a binder and winter nights fall well below freezing.

Where PG 76-16 is the wrong answer even inland

• Below the wearing course. The usual structure places the modified binder only where shear stress is highest — the surface course at roundabouts, bus lanes, port approaches and signalised intersections — with an unmodified grade such as PG 64-16 in the layers beneath, which see far less shear and do not justify modification. Specifying PG 76-16 through the full depth is a common habit that adds handling obligations without adding performance below the surface course.
• Where reclaimed asphalt content is high. Every tonne of reclaimed material carries oxidised binder into the blend and drags the effective low grade warmer. With only six degrees between −16 and −10, the RAP fraction and the virgin binder grade have to be settled in the same conversation rather than one after the other.
• Where destination storage cannot hold and agitate a modified binder. This is a real constraint and not a technicality. A site with an unagitated tank and no reliable temperature control will not keep a modified binder homogeneous, and the material that comes out of the bottom of that tank at the end of a month is not the material that was tested at the load port.

M320 sets two handling numbers, and neither is an operating temperature

The flash point minimum of 230 °C by ASTM D92 is a safety ceiling. The rotational viscosity limit of 3 Pa·s at 135 °C by AASHTO T316 establishes that the binder can be pumped out of a tank and will coat aggregate. Neither is a storage temperature and neither is a mixing temperature, and treating either as one is a common and avoidable error.

The flash point is a boundary, not a target

The 230 °C minimum is measured on the Cleveland open cup, ASTM D92 / AASHTO T48, which is the correct apparatus for a paving grade. The Tag open cup methods — ASTM D1310 and ASTM D3143 — belong to cutback asphalts, whose solvent gives them far lower flash points, and a Tag result printed on a performance grade certificate is a safety-case error rather than a clerical one. Work to the measured flash point of the batch on the certificate, not to the specification minimum, and keep every working temperature well below it.

Never apply a direct flame to a tank, a drum, a valve or a line. Open-flame burners, blow torches, improvised fires under drums and unshielded heating elements are the direct cause of bitumen fires and of the burns that go with them; heating must be by thermal oil, steam, or correctly rated and controlled electric elements that are fully immersed. Keep tanks, drum stacks and vents clear of naked flames, welding and cutting work, hot exhausts and other ignition sources, and treat any repair to a vessel that has held binder as hot work requiring a permit. Keep water out of a hot tank or drum: trapped water flashes to steam beneath the binder and can eject the contents violently. Hot binder causes severe burns that adhere to skin, so face protection, gauntlets and covered arms belong at every sampling point and every drum opening. A reference to the safety data sheet supplements a correct heating and storage procedure; it does not replace one.

The viscosity line deserves particular attention on this grade. M320 permits the requirement to be waived at the discretion of the specifying agency where the supplier warrants that the binder can be adequately pumped and mixed at temperatures meeting all applicable safety standards, and that provision exists precisely because some modified binders exceed 3 Pa·s at 135 °C. A documented waiver with a supplier warranty behind it is a legitimate position. A certificate with no viscosity result and no statement about it is a different thing entirely, because the buyer then has no information at all about whether the material will move through his pump.

The equiviscous method does not apply

For an unmodified binder a plant reads its temperatures off the binder’s own viscosity–temperature chart at the conventional equiviscous targets, near 0.17 Pa·s for mixing and 0.28 Pa·s for compaction. Those two viscosity figures are long-standing laboratory practice rather than requirements of any standard — M320 sets no mixing or compaction temperature at all — and they are quoted here only to explain the method the plant is used to. That method rests on the binder behaving as a simple Newtonian liquid at plant temperatures, plotting as a straight line on the chart. A polymer modified binder does not. Its viscosity is shear-rate dependent and the polymer’s contribution does not disappear neatly with heat, so the equiviscous construction returns temperatures that are too high — and running a modified binder too hot is the single most effective way to destroy the property that was paid for.

The supplier’s recommended mixing and compaction temperature range replaces the calculation. Ask for it in writing with the offer rather than after the cargo lands, and give it to the plant with the delivery documents. The same applies to the compaction window: modified mixes generally stiffen faster as they cool, so roller patterns and haul distances that worked with a neat binder may not transfer.

Separation is a documented behaviour, not a theoretical risk

A polymer modified binder is a two-phase system. The polymer-rich phase is swollen with the maltene fraction and is less dense than the asphaltene-rich phase around it, so a binder held hot and static will tend to segregate, with the polymer-rich material migrating upward. The consequence at destination is that the top of a tank grades differently from the bottom, and the difference grows with time and temperature.

ASTM D7173 is how that tendency is measured: a sample sealed in a tube, held vertically at 163 °C for 48 hours, cooled, cut, and the top and bottom sections recovered and tested separately, with the difference reported — commonly as a difference in softening point by ASTM D36, sometimes as a difference in a DSR result. Read the method for what it is. D7173 produces a number; it does not set a pass mark. The acceptance limit is written into the purchase specification, and a supplier who names the method without a limit has quoted half a requirement.

What follows operationally is unglamorous and effective. Keep the binder circulating, either continuously or on a defined schedule, rather than letting a hot tank stand. Draw from a circulating tank rather than from a dead leg. Avoid leaving a part-full tank hot through a shutdown. Where material has stood static and hot for an extended period, treat re-testing as the default position rather than assuming the load-port certificate still describes it.

Temperature and time both degrade a polymer network

Prolonged storage at high temperature degrades an elastomeric network. The practical signature is a loss of elastic response before anything else — percent recovery falls, then the high-temperature grade follows — and it is not recovered by adding heat or by stirring harder. No standard states the temperature or the residence time at which this becomes significant, because both depend on the polymer, the dosage and the base binder. The supplier’s technical data sheet governs, and it should arrive as part of the offer documentation rather than being requested after the fact.

Two habits cover most of the risk. Heat what will be used rather than holding the whole inventory at working temperature, and log tank temperatures so that a long hot residence is visible as a record rather than remembered as an impression.

Bulk, drums and bags

• Bulk and tank containers are the least damaging route for a modified binder where the destination has heated, agitated storage, because the material is heated once at loading and once at discharge rather than being taken from cold to working temperature twice.
• New steel drums are shipped and used routinely for modified grades, and the reheating step is where the damage happens. A drum oven or hot room brings the outside of a drum to temperature long before the middle, and a local overheat at the drum wall is exactly the condition that degrades a polymer network. Where drums are unavoidable, agree the reheating equipment and procedure before shipment rather than discovering the plant’s method after arrival, and specify new drums with sound closures — a reconditioned drum is the usual source of both contamination disputes and leaks.
• Meltable and poly bags are attractive because the package goes into the mixer with the binder, removing the decanting step altogether. The film is then part of the mix, so its compatibility with the plant and with the project specification has to be confirmed in advance rather than assumed.

Loading arithmetic for a 20' container

The planning figures commonly used in the trade for a paving grade shipped in a 20' FCL are set out below. They are commercial practice, not requirements of any standard, and the drum count depends on the drum dimensions and the stow actually used:

• 150 kg new steel drums — 80 drums per container, 12 MT net
• 180 kg new steel drums — 80 drums per container, 14.4 MT net
• 185 kg new steel drums — 80 drums per container, 14.8 MT net
• 1 MT jumbo or poly bags — 20 bags per container, 20 MT net

Three qualifications belong with those figures. They are net cargo weights, so gross weight including drums or bags is higher and it is gross weight that meets the container’s rated payload and the destination’s road weight limits. The rated payload of the container and the carrier’s own limits govern in the end, and they are not identical everywhere. And a paving grade is contracted by mass, unlike a cutback or an emulsion, which are normally sold by volume — so no volume conversion enters a PG 76-16 order except at the destination tank, where capacity has to be checked at the binder’s density at storage temperature rather than at 15 °C.

What to require on the certificate

A performance grading report is batch-specific and does not travel. Production graded last quarter says nothing enforceable about the tank being loaded against your contract, and on a modified product the polymer content and dispersion are batch properties as much as the grade is. Ask for the following against the tank or batch reference that appears on your Certificate of Analysis:

• Tank or batch number, sampling date and test date on every page.
• The full M320 set at the derived temperatures: DSR at 76 °C on original binder and on RTFOT residue, fatigue at 34 °C on PAV residue, bending beam at −6 °C with both S and the m-value.
• The ageing conditions printed rather than implied: RTFOT to AASHTO T240 at 163 °C for 85 minutes, and PAV to AASHTO R28 for 20 hours at 2.1 MPa with the temperature stated.
• Continuous grade at both ends, not only the rounded designation, because that is what shows whether the binder carries margin inside a six-degree envelope.
• A statement of whether the binder is modified and with which family of modifier.
• A separation result to ASTM D7173 against a limit named in the contract.
• Where AASHTO M332 governs, MSCR to AASHTO T350 on RTFOT residue at the high-temperature grade printed in the designation — 76 °C for a PG 76V-16, but 64 °C where the same job is written PG 64V-16 — reporting Jnr at both stress levels, the Jnr difference and percent recovery.
• The measured flash point for the batch, and the supplier’s recommended storage, mixing and compaction temperature ranges.

None of that is an unusual request for a modified performance grade, and a supplier who routinely produces one will have most of it on file. The pattern worth noticing is not a missing line here or there but a certificate that answers the easy questions in detail and the decisive ones not at all.